Report Colluvium

LANDSLIDES OF THE CINCINNATI, OHIO, AREA
Landslides in
Colluvium
U.S. GEOLOGICAL SURVEY BULLETIN 2059-B
Landslides in Colluvium
By ROBERT W. FLEMING and ARVID M. JOHNSON
U.S. GEOLOGICAL SURVEY BULLETIN 2059-B
The behavior ofcolluvial
landslides is strongly affected
by differences in thickness
UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 1994
U.S. DEPARTMENT OF THE INTERIOR
BRUCE BABBITT, Secretary
U.S. GEOLOGICAL SURVEY
Robert M. Hirsch, Acting Director
For sale by
U.S. Geological Survey, Map Distribution
Box 25286, MS 306, Federal Center
Denver, CO 80225
Any use of trade, product, or firm names in this publication is for descriptive purposes only and
does not imply endorsement by the U.S. Government.
Library of Congress Cataloging-in-Publication Data
Reming, Robert W.
Landslides in colluvium; the behavior of colluvial landslides is strongly affected by
differences in thickness / by Robert W. Reming and Arvid M. Johnson.
p. cm. (U.S. Geological Survey bulletin ; 2059)
(Landslides of the Cincinnati, Ohio, area; ch. B)
Includes bibliographical references.
Supt. of Docs, no.: 119.3: 2059
1. Landslides Ohio Cincinnati Region. 2. Colluvium Ohio
Cincinnati Region. I. Johnson, Arvid M. II. Title. III. Series.
IV. Series: Landslides of the Cincinnati, Ohio, area ; ch. B.
QE75.B9 no. 2059
[QE599.U5]
557.3 s dc20 93-15779
[551.3'07] CIP
CONTENTS
Abstract................................................................................................................................. Bl
Introduction.......................................................................................................................... 2
Acknowledgments ..................................................................................................... 2
Geology and Physiographic Setting ..................................................................................... 2
Delhi Pike Landslide Area ................................................................................................... 3
Delhi Pike Study Area ......................................................................................................... 5
Thick Landslides ........................................................................................................ 5
Sequence of Movement ....................................................................................... 7
Deformation Caused by Thick Landslides .......................................................... 10
Subsurface Water in Thick Landslides ................................................................ 12
Summary of Thick Landslides ............................................................................. 13
Thin Landslides ......................................................................................................... 15
The Thin Landslide at Delhi Pike ........................................................................ 16
Trenches and the Failure Surfaces ....................................................................... 19
Properties of the Colluvium ................................................................................................. 20
Stability Analyses ................................................................................................................ 22
Concluding Remarks ........................................................................................................... 23
References Cited .................................................................................................................. 24
PLATE
[Plate is in pocket]
1. Profiles of trenches and map of landslide features, borings, and trenches in the Delhi Pike study area near
Cincinnati, Ohio.
FIGURES
1. Map showing location of the Delhi Pike study area and important geographic features ......................................... B3
2. Geologic map of the Delhi Pike study area .............................................................................................................. 4
3. Photographs showing emergent toes of deep-seated landslides near intersection of
Hillside Avenue and Darby Road ............................................................................................................................. 6
4. Photograph of landslide scarp for thick slide that heads on Delhi Pike .................................................................... 7
5. Sketch map of positions of tilted pine trees, walls, and landslide toes on the property
at 5434 Hillside Avenue ........................................................................................................................................... 8
6. Photographs of two pine trees tilted by landslide movement at 5434 Hillside Avenue ........................................... 9
7. Graph showing thickness of growth rings in the "west tree" of figure 6, plotted as a
function of time ......................................................................................................................................................... 10
8. Graph showing thickness of growth rings in the "east tree" of figure 6, plotted as a
function of time ......................................................................................................................................................... 11
9. Photograph of bulge produced by landslide toe in asphalt parking area at Anderson
Ferry Road and River Road....................................................................................................................................... 12
10. Photograph of landslide toe in Mount Adams section of Cincinnati ........................................................................ 13
11. Photograph of near-vertical bump in landslide toe ................................................................................................... 13
12. Water-level record of 1 year for an open-tube piezometer in a limestone layer in
bedrock .........................................._^ 14
m
IV CONTENTS
13. Photograph looking uphill along Delhi Pike from the location of trench 2 .............................................................. B15
14. Photographs of thin landslide along Columbia Parkway, east of downtown Cincinnati.......................................... 17
15. Water-level and precipitation record for an open-tube piezometer placed at contact between colluvium and
bedrock..................................................................................................................................................................... 18
16. Scanning-electron micrographs of portions of the failure surface ............................................................................ 20
17. Photograph of kinked roots on an exposed failure surface ....................................................................................... 20
18. Graph of hypothetical force-ratio-displacement models for thin and thick landslides ............................................. 23
TABLES
1. Summary of selected physical properties of Eden and Pate Soil Series ................................................................... B21
2. Summary of physical properties of colluvium and shale from the Delhi Pike study area......................................... 21
3. Residual strength of colluvium from the Delhi Pike study area ............................................................................... 22
CONVERSION FACTORS
For the convenience of readers, the metric units used in this report may be converted to inch-pound units by using the following
factors:
Multiply metric units By To obtain inch-pound units
micrometers (um) 3.937xlO~5 inches
millimeters (mm) 0.03937 inches
centimeters (cm) 0.3937 inches
meters (m) 3.281 feet
kilopascals (kPa) 0.145 pounds per square inch
LANDSLIDES IN COLLUVIUM
i f
By Robert W. Fleming and Arvid M. Johnson
ABSTRACT
The most common and destructive landslides in the
Cincinnati, Ohio, area are formed in colluvium. Colluvium is
formed from bedrock units by weathering and slaking of
shales. The principal colluviurn-producing bedrock unit is
the Kope Formation, which occupies the lower 60-70 m of
hillslopes between the level of the Ohio River and an upland
some 150 m above the river level. This formation, which
contains about 80 percent shale and 20 percent limestone,
slakes readily to produce a stony, silty clay colluvium. Over-
lying formations contain smaller amounts of shale, typically
produce smaller amounts of a more stone-rich colluvium,
and support steeper slopes. The colluvium forms a wedge-
shaped mass ranging up to about 15 m in thickness. In the
upper parts of the slopes, grades are steeper and the collu-
vium is thinner.
We recognize a significant difference in landslide
behavior in slopes underlain by thick and thin colluvium.
Landslides in thick colluvium that is, more than 2 m
thick typically occur in the spring but can occur at any time
of the year in response to a disturbance, such as grading.
During a movement episode, the landslides typically move
only a few centimeters to perhaps a meter. In form, these
thick landslides consist of a single zone of overlapping
scarps at the head and multiple toe bulges downslope. The
flanks of movement are indistinct; shear displacement is evi-
dently distributed over a broad zone. Analysis of a 37-yr
record of movement, as revealed in disturbance to trees at
our Delhi Pike study site, showed a small amount of move-
ment nearly every year. During the 37-yr interval, there were
two episodes of abrupt movement of several centimeters,
both during years of above-average precipitation. For the
most part, however, movements occurred during years of
normal precipitation, when water levels were below the fail-
ure surface except at the most downslope toe of movement.
'U.S. Geological Survey, Box 25046, MS 966, Denver, CO 80225.
'lDept. of Earth and Atmospheric Sciences, Purdue University, West
Lafayette, IN 47907.
Landslides in thin colluvium that is, less than about 2
m thick typically occur in the spring after the ground has
thawed and before the vegetation has fully blossomed. Their
movements are associated with rainfall, but precise timing
has not been measured. Measurements of water levels in thin
colluvium reveal that, during most springtimes, the collu-
vium is saturated to the ground surface for brief intervals.
The thin landslides take the form of multiple scarps and a
single toe of movement. For most thin landslides, the flanks
are well-defined, simple boundaries. The multiple scarps are
evidence for stretching or extension being the dominant form
of kinematic behavior. Indeed, thin landslides are character-
ized by stick-slip behavior in which the landslides may
accelerate after initial failure and slide completely out of
their scars. During movement, they commonly pull them-
selves apart, leaving deposits that consist of separated hum-
mocks or slabs of colluvium.
Four trenches were dug into the colluvium to examine
the failure surfaces and to obtain samples for laboratory test-
ing. The failure surfaces are paper thin, highly polished, and
striated. At all scales of observation, roughness is evident in
the surfaces. At a scale of a few micrometers, the surfaces
have small steps between shiny surfaces that appear to be
analogous to chatter marks on rocks in fault zones. At a
larger scale, the surfaces are scratched by fossil fragments
and sand-size grains in the colluvium. At a still larger scale,
the surfaces bifurcate around rock fragments, and multiple
surfaces occur where the shape of the failure surface is
changing.
Residual shear tests were conducted on samples from
near the failure surfaces. Strength parameters varied depend-
ing on the way the sample was prepared and tested. Mini-
mum values were obtained in a ringshear device on samples
that had the coarse fractions removed. A reasonable estimate
of the residual strength parameters is an angle of internal
friction of 16 degrees with a small cohesion intercept of 10
kPa. Other typical properties of the colluvium are LL=45,
PI=22, sand =10 percent, silt = 30 percent, and clay = 60 per-
cent. The clay minerals are predominantly illite and mixed-
layer clay. Stability analyses conducted by back-calculation
gave expectable results for the thick landslides. At equilib-
rium, our measured values of residual strength are adequate
Bl
B2 LANDSLIDES OF THE CINCINNATI, OHIO, AREA
to support the slope for the geometry and water conditions
observed. For the thin landslides, the measured strength
appears insufficient to support the slope. Additional strength
provided by tree roots, roughness of the failure surface, or
small residual cohesion could account for the discrepancy
between apparent strength and stability.
INTRODUCTION
Annual per capita costs of damage due to landsliding in
Hamilton County, Ohio (Cincinnati and vicinity), are among
the highest of anywhere in the United States. Annual per
capita costs for the period 1973-78 were $5.80, unadjusted
for inflation; total costs of damage were nearly $31 million
for the same 6-year period. Not included in this total was
nearly $30 million expended after 1978 for a single landslide
that occurred in 1974 in the Mt. Adams section of Cincinnati
(fig. 1) (Fleming and Taylor, 1980).
A separate study of costs of landslide repair, done by
students and faculty from the University of Cincinnati, found
that the annual direct cost of emergency repairs to local
streets in the City of Cincinnati is about $0.5 million.
Deferred repairs of landslide damage to Cincinnati streets
amounted to about $18.5 million in 1987 (Earth Surface Pro-
cess Group, 1987).
The most common and destructive landslides are in
slopes underlain by colluvium. The colluvium is derived by
weathering of subjacent bedrock and accumulates on slopes
as a wedge-shaped mass of stony clay.
Two distinct types of landslides occur in the colluvium.
The two types have different morphologies and kinematic
behavior, and they appear to be triggered by different stim-
uli. The most striking difference between the two types of
landslides is in thickness. Thin landslides, less than 2 m
thick, occupy the upper parts of hillslopes where the collu-
vium is thin. Thick landslides, more than 2 m thick, occupy
the lower parts of the slopes.
In this chapter, we describe the Delhi Pike area, which
is an area of landsliding we believe to be representative of
landslides in colluvium throughout the Cincinnati metropol-
itan area. The area has been the focus of several detailed
studies reported in other chapters in this bulletin. The pur-
pose of this chapter is to describe the physical setting and
movement styles of the landslides and, thus, to provide a
context for the more narrowly focused chapters that follow.
ACKNOWLEDGMENTS
Onder Gokce, University of Cincinnati, helped map the
landslides, logged borings and trenches, and read inclinom-
eters and piezometers. He incorporated many of those data in
his Ph.D. dissertation. Sherry Agard, U.S. Geological Sur-
vey, Denver, analyzed the tree cores. Roger Nichols, U.S.
Geological Survey, Denver, conducted the drilling and
testing program. The base map was prepared from aerial
photographs by James Derrick. Property owners Dale
Schmale, Edgar Alien, and the College of Mount St. Joseph
granted access for the study. To all these people and to the
college, we express our sincere appreciation.
GEOLOGY AND
PHYSIOGRAPHIC SETTING
The landscape in the Cincinnati area consists of a dis-
sected upland surface, hillslopes along the Ohio River and
principal tributaries, and flood plains and terraces. Maxi-
mum relief is about 150 m. The uplands are mantled mainly
by glacial deposits, mostly of Illinoian age. Sedimentary
bedrock underlies the till at various depths. Bedrock is typi-
cally exposed in the upper parts of hillslopes as ledges of
limestone and in the floors of some of the smaller postglacial
valleys. The Ohio River and its tributaries (Mill Creek, Lick-
ing River, Little Miami River, Miami River, and Whitewater
River; fig. 1) have been strongly imprinted by their glacial
history and have flood plains and terraces of varying widths.
Materials in the terraces and flood plains vary from sand and
gravel to laminated silts and clays.
The bedrock geology of the Cincinnati area consists of
very gently dipping (1-2 m/km) shale and limestone of Late
Ordovician age. The shale beds in the bedrock sequence are
poorly cemented and slake to their constituent grains in
response to moisture and temperature changes. Three of the
bedrock units contain significant amounts of shale and pro-
duce colluvium. These are, from the oldest, the Kope Forma-
tion, the Fairview Formation, and the Miamitown Shale. The
Kope Formation occurs near the level of the Ohio River and
occupies the lower 60-70 m of hillslopes along the Ohio
River and its tributaries. The overlying 30-m-thick Fairview
Formation and undifferentiated bedrock as much as about 50
m thick occupy the remaining slope between the river and an
upland area, which is typically mantled by loess or glacial
deposits.
In general, the amount of shale decreases upsection.
The lower part of the Kope Formation contains about 80 per-
cent shale. Shale content diminishes to about 70 percent in
the upper 12m of the Kope Formation, and the Fairview For-
mation and overlying undifferentiated bedrock contain less
than 50 percent shale (Osborne, 1974; Ford, 1974; Luft,
1971,1972; and Gibbons, 1972). The changing shale content
commonly is expressed as a slight break in slope at the con-
tact between the Kope and overlying formations; slopes
formed on the Kope Formation are a few degrees flatter than
slopes on overlying bedrock.
The colluvium varies in thickness depending on its
position on the slope. In the lower part of the slope, collu-
vium is as thick as about 15 m. Colluvium thickness
diminishes to a meter or less near the top of the Fairview For-
mation in the upper part of the slope.
8430'
B3
3915'
3900
0 5 10 MILES
0 5 10 KILOMETERS
Figure 1 . Location of the Delhi Pike study area and important geographic features in the Cincinnati, Ohio, area.
DELHI PIKE LANDSLIDE AREA
The Delhi Pike landslide area is about 12 km west of
downtown Cincinnati in unincorporated Hamilton County,
Ohio (fig. 1). The Delhi Pike landslides are just a few of the
many landslides that occur within a narrow stretch of the
Ohio River Valley extending from Mill Creek to the Miami
River about 22 km farther west. This stretch of the valley
was created when glacial ice of Illinoian age blocked a more
northern channel, creating a lake. Silts and clays deposited in
the lake have been involved in major landslides in other parts
of the metropolitan area. The spilling of the lake eroded the
modem channel of the Ohio River (Durrell, 1961a, 1961b).
The total width of this narrow section of the valley, as
measured between the uplands of Ohio and Kentucky, is
about 1.5 km, compared to a width of 2.5 km upstream and
4.0 km downstream. The narrow part of the valley lacks a
continuous flood plain or terraces and essentially consists of
slopes of colluvium from the level of the Ohio River to
ledges of limestone outcrop near the top. Landslides have
occurred along these slopes on both sides of the Ohio River.
In 1973, extensive landsliding resulted in temporary closure
of a 5-km stretch of State Route 8 west of Bromley, Ky. (fig.
1), and in the destruction of several homes. At the same time,
in Ohio, landslides were destroying homes and sections of
several roads. As part of this widespread landslide disaster,
multiple landslides caused the permanent closure of Delhi
Pike in our study area (The [Delhi Township] News, April
11,1973, p. 1, "Sister's Hill too hazardous; closed by County
Engineer").
Many of the streets, buildings, and railroads in the
Delhi Pike area were constructed before the turn of the cen-
tury. A topographic map of Cincinnati published in 1912
shows the same street and railroad alignments in the area that
exist today. Transportation routes are parallel to the Ohio
River and trend along the hillslopes roughly parallel to topo-
graphic contours. From the level of the Ohio River, at pro-
gressively higher elevations, are railroad tracks (12 m),
River Road (about 20 m), and Hillside Avenue (about 40 m
upslope from the river) (fig. 2). Delhi Pike was an old road
that provided a connection between Hillside Avenue at ele-
vation 170 m and Delhi Township at elevation 260 m in the
uplands. A few homes were built on the downslope side of
the lower part of Delhi Pike, but the upslope side was appar-
ently too steep for development.
The colluvial slope has an overall concave-upward pro-
file. Slope inclination increases from about 8 or less near
the Ohio River to about 10-12 along the section between
River Road and Hillside Avenue. On the upslope side of
Delhi Pike, the inclination is 18-25. The initial character
of the lower, more gently sloping portions of the slope has
been obscured by development. In undeveloped areas in
other parts of the county, slopes at this level have a smooth
but lobate texture; the lobes, typically 1 or 2 m high and
10-20 m across, appear to represent the distal ends of land-
slides or earthflows.
Only a few landslides have been active in historic time
in the gentle slopes along River Road. Two notable slides
occurred at North Bend, about 10 km downstream from
Delhi Pike, and near Riverside-Harrison School, about 8.7
km upstream. The landslide at North Bend, near the William
Henry Harrison Monument, occurred during the 1970's
above the outside of a bend in the river, where the slope was
locally oversteepened by river erosion and loaded with a
small amount of fill. The other landslide, near the Riverside-
Harrison School, occurred in response to an excavation for a
railroad roundhouse in 1927 and ultimately destroyed about
40 homes (Von Schlichten, 1935). Both examples demon-
strate that the thick colluvium near the level of the Ohio
River will fail by landsliding if the hillslope is adversely
disturbed.
Within the Delhi Pike study area, the slopes have failed
extensively (fig. 2). Hillside Avenue is deformed along
much of its length, and numerous empty lots containing
Figure 2 (above and facing page). Geologic map of the Delhi Pike study area. See figure 1 for location. Contact between the Kope (Ok)
and Fairview (Of) Formations is exposed in cut slope in western part of the map area. Contact was projected through map area using apparent
dip angle from published geologic maps of areas directly across the Ohio River in Kentucky (Gibbons, 1972). The Fairview Formation was
assumed to be 30 m thick.
LANDSLIDES IN COLLUVIUM B5
house foundations and overgrown remnants of driveways
and walls attest to a long history of slope movements. Still
farther upslope are scarps, scars, and lobes produced by fail-
ure of a thinner mantle of colluvium.
DELHI PIKE STUDY AREA
Within the area of extensive landslides shown in fig-
ure 2, a smaller area containing both thick and thin land-
slides was selected for detailed study (pi. 1). The trace of
Delhi Pike extends east-west across the middle of the area
and divides shallow landslides uphill from deep-seated
landslides downhill. Subsurface conditions were investi-
gated with borings and trenches. Instrumentation was
installed in both shallow and deep slides to determine the
geometry and water conditions and to obtain samples for
laboratory testing.
THICK LANDSLIDES
The heads of the thick landslides are in the fill on the
downhill edge of Delhi Pike. The fill contains overlapping
low scarps and tension cracks (pi. 1) extending to the break
in slope at the shoulder of the road. These cracks, which indi-
cate stretching or extension of the landslide material, are
confined to the fill area. All indications of landsliding farther
downslope are of shortening or compression. The compres-
sional features are difficult to recognize in natural slopes but
are well expressed where they intersect a wall, road, or walk-
way. Figure 3 consists of two photographs of toes of deep-
seated landslides; locations are shown on the map of the
Delhi Pike landslide complex (pi. 1).
The lowermost toe (fig. 3A) is a broad bulge in Hillside
Avenue, where the vertical component of displacement is
larger than the horizontal component. The edge of the road
showed no visible offset even though the bump grew percep-
tibly over the 5-yr period of observation. Farther uphill, dis-
placement on another toe of movement (fig. 3B) is almost
entirely horizontal.
The deep-seated landslide did not displace inclinometer
casings enough during our monitoring period of 1979-80 to
allow us to accurately determine the positions of failure sur-
faces. Based on borings, the maximum depth to the failure
surface was about 4 m at the head of the slide (including the
thickness of side-hill fill), 2.3 m near the stone walls (boring
Wl), about 6 m near boring 9, and about 3.5 m at boring 20
(locations shown in pi. 1).
The shape of the failure surfaces can be inferred from
the mappable features at the ground surface. The head of
the landslide consists of a narrow zone, as much as 4-5 m
wide, of extensional faults and cracks. The faults and
cracks are vertical to steeply dipping downslope. The hori-
zontal offset of Delhi Pike is smaller than the vertical sepa-
ration of blocks of displaced road (fig. 4). Bulges and other
indications of shortening are evident at the first break in
slope downslope from the head. These bulges apparently
20 0 40 0 60 0 80 0
10 0 20 0
C ON TOU R IN TERV A L 25 F T (7.6 M)
10 0 0 F EET
V
30 0 METERS
Undifferentiated
Fairview Formation
Kope Formation
Upper
Ordovician
EXPLANATION
Bedrock contact Dashed where approximate
Landslide scarp Dashed where approximate
Landslide toe Dashed where approximate
Strike-slip fault Showing direction of movement
B6
Figure 3. Emergent toes of deep-seated landslides near intersection of Hillside Avenue and Darby Road. See map on plate 1 for locations.
A, View west toward the intersection. The broad bump in the middle of view is a toe for a thick landslide that extends 60 m upslope. Note
the tilted utility pole at the left of the view and damage to the stone wall on the right. The asphalt paving over the bump has recently been
replaced. The dark-colored marks in the eastbound (oncoming) lane are produced by vehicles scraping on the road surface after bouncing
over the bump. The road has not been significantly displaced laterally; the principal movement direction is vertical. B, Stairway, about 18
m farther upslope, displaced a few centimeters to the right (south). Principal movement direction is horizontal.
LANDSLIDES INCOLLUVIUM B7
Figure 4. View to northeast of landslide scarp for thick slide that
heads on Delhi Pike.
reflect predominantly horizontal displacement; a cistern
that was originally about 3 m from the stone wall (Mr. Dale
Schmale, oral commun., 1980) near boring 14 is now only
2 m away, and the wall has not been perceptibly displaced
vertically.
The next series of bulges and thrusts occurs just down-
hill from a break in slope above Hillside Avenue (pi. 1, fig.
3). The displacement of the stone and concrete work on the
slope is nearly horizontal. Still farther downslope, at Hill-
side Avenue, the displacement direction is predominantly
upward. This toe of movement appears to be the downslope
limit of active sliding. In plan view, the toes are lobate and
have local relief of a few centimeters to about 2 m.
Trench 1 was placed through one of the landslide
lobes on the east side of the study area. Here, the colluvium
consists of two separable units, a brown pedologic soil
overlying yellowish-gray colluvium (pi. 1). One slicken-
sided failure surface was found in the trench wall. Toward
the upslope end of the trench, the surface was within the
yellowish-gray colluvium; toward the downslope end of the
lobe, the yellow-gray colluvium had overridden part of the
brown pedologic soil. The failure surface was essentially
horizontal at a depth of about 2 m. Boring 11 is about 4 m
west of trench 1 and contains an inclinometer casing. The
casing, originally 6.6 m long, has been constricted at a
depth of 3.6 m (measured May 10, 1990). Thus, at this lobe
of the landslide complex, at least two failure surfaces exist,
one at a depth of about 2 m and the other at 3.6 m. The fail-
ure surface at 2 m apparently created the lobe containing
trench 1 and boring 11 (pi. 1). The failure surface at 3.6 m
may reach the surface at the 560-ft (171-m) contour. If so,
the deeper failure surface slopes about 7 downhill toward
Hillside Avenue. We do not know whether movement has
occurred on the shallower surface since our study began,
but if it has, the amount was insufficient to constrict the
inclinometer casing at that depth.
SEQUENCE OF MOVEMENT
A sketchy history of the overall movement of the thick
landslide can be inferred from deformation of features on the
landslide and from an interview with the property owner,
Mr. Dale Schmale (oral commun., 1981). Prior to abrupt
movement of the landslide in the spring of 1973, the property
identified as 5434 Hillside Avenue contained four struc-
tures a house, a barn, a garage, and a shed. The house was
at least 100 yr old and was situated on the flat area near bor-
ings 9, 10, and 14 (pi. 1). The garage, barn, and shed were
east of the house near trench 1. Mr. Schmale purchased the
property in 1972 and was involved in remodeling the house
when it was destroyed by movement. The first indication of
trouble was when a 400-yr-old tree, located at a toe of move-
ment just east of boring Wl, toppled onto the house
(Schmale, oral commun., 1981). At that time, there were
cracks in Delhi Pike, but Mr. Schmale had noticed no evi-
dence for sliding on his property. Within a few days after the
tree toppled, the house and other structures were deformed
beyond repair. The entire episode of movement occurred in
a 3-week period of April 1973.
As recently as 1990, the overall appearance of the prop-
erty remained as it was in 1973. Small amounts of movement
had occurred that were sufficient to squeeze the borings
closed and to reestablish the bulge in Hillside Avenue, in
spite of repeated repairs. However, there had been no epi-
sodes of abrupt movement similar to the episode in 1973.
Additional inferences can be made about movement of
the thick landslides from disturbance to trees. Trees on
undisturbed hillslopes tend to grow vertically from the tip of
the stem and to form concentric growth rings (Kozlowski,
1971). When a tree is tilted, it responds to the change both
internally and externally. New growth from the stem tip con-
tinues to grow vertically, producing a bend or curve in the
stem. Eccentric growth in the trunk creates asymmetric tree
rings and reaction wood. In conifers, the growth rings are
wider and reaction wood forms on the downslope side
(Agard, 1979). In general, reaction wood can be distin-
guished visually from normal wood in that it is darker, has a
denser, smoother appearance, and has little contrast between
early wood and late wood in a growth ring (Agard, 1979).
Tree cores were used to examine the growth behavior of
two pine trees on the thick landslide at Delhi Pike. The pine
trees were planted in the early 1940's in an open area along
the driveway to the house at 5434 Hillside Avenue. Both
trees have been tilted and deformed in response to landslide
movements. The positions of the trees relative to nearby
landslide features are shown in figure 5.
Figure 6A is a photograph of the two pine trees viewed
northwest from near the site of the house at 5434 Hillside
Avenue. The tree on the left is termed the "west tree" and
the tree on the right the "east tree." The bulge in the fore-
ground is a toe bulge that, from right to left, trends through
dry grass and crosses the ivy-covered stone wall about in line
with the west tree. The bulge passes just to the right of the
Photo point
fig. 6B
N
(Approx.)
20 METERS
+ + +
W-2
EXPLANATION
Topographic contour Contour interval 2 ft (0 6 m)
Landslide toe
Core number and direction
Figure 5. Sketch map of positions of tilted pine trees, walls, and landslide toes on the property at 5434 Hillside Avenue.
west tree, and landslide movement has pushed the tree
toward the left or downslope. In response, the tree has
grown straight for about 5 m and then is continuously curved
except for one small, more abrupt bend about halfway to the
top. The continuously curved portion of the tree appears to
be a response to small annual increments of landslide
movement, and the single, abrupt bend could be a response
to one event of more than normal movement of the landslide.
The east tree is about midway between two toe bulges
(fig. 5). The same toe bulge that is pushing the west tree
forward is about 3 m downslope from the east tree. Another
toe bulge is about 3 m upslope from the east tree. The trunk
of the east tree is straight in the lower 5 m and has a gentle
bend over the next 5 m to an abrupt bend about 10 m from
the base (fig. 6B). The abrupt bend is apparently a response
to a dramatic uphill tilt to the tree. In the year following the
abrupt tilting, the east tree put out a large branch on its uphill
side (fig. 6fi). From the abrupt bend to the top of the tree, the
trunk is somewhat, but not systematically, contorted. Count-
ing sets of branches from the tops of both trees down to the
abrupt bends shows that both bends were created about the
same time, a little more than 20 yr before the 1980 growing
season.
Cores were taken from the trees at heights of 30-50 cm
above the ground. The core numbers and directions of coring
are shown on figure 5. The cores were examined microscop-
ically and the rings were counted and measured by Sherry
Agard (U.S. Geological Survey, Denver, Colo.).
The results for the west tree are in figure 7. Shown are
the thicknesses of the growth rings for the downhill (W3),
uphill (Wl), and the lateral (W2 and W4) sides of the tree.
Also plotted is the percent eccentricity of the growth rings on
opposite sides of the tree. Both W2 and W4 penetrated the
center of the tree and established that the tree was a 40-cm-
high sapling in 1943.
For W2 and W4, the growth rings are almost exactly the
same size through 1947, when the east side of the tree (W4)
began producing consistently thicker rings. This pattern
continued through 1980 with exceptions only in 1962-63
and 1970. For cores W3 and Wl, the downhill portion of the
growth ring, which is the direction of tilt, was consistently
thicker than the uphill portion. Exceptions occurred only
during 1963-65 and 1969.
The data for the east tree are plotted on figure 8. Cores
E3 and El are the downhill and uphill sides of the tree,
respectively, and E4 and E2 are the left and right sides of the
tree as viewed looking uphill. E6 is a core taken from a point
just above the abrupt bend about 10m above the ground sur-
face. The E6 core passed through the center of the trunk and
continued through the tree.
Figure 6. Two pine trees tilted by landslide movement at 5434
Hillside Avenue. A, Both trees as seen from the southeast. "West
tree" is on left and "east tree" is on right. B, "East tree" as seen from
the northwest
The east tree is strongly tilted upslope, and the El
growth rings (uphill side) are consistently thicker than those
on the downhill side. Exceptions are in 1949-50, 1955,
1962-63, and 1968. The trend of the eccentricity shows
increasing thickness of the uphill side of the ring relative to
the downhill side. The core for E6 dates the year of strong
tilting as 1958 or 1959.
For both the east and west trees, the year 1959 is a year
of markedly suppressed growth. The east tree also experi-
enced suppressed growth in 1958. Growth was somewhat
suppressed in both trees in 1965, but we have found no other
evidence of landslide movements that year.
The combined observations of external changes to the
trees (curved trunks and abrupt bends) and the tree-ring data
lead to the following conclusions. There were active slope
movements at the site as early as the mid-1940's, but the
tree-ring data cannot be extended back in time beyond those
years. The most dramatic movement apparently occurred in
1958, following near-record precipitation in calendar year
1957. Abrupt tilting and suppressed growth affected both
trees during the 1959 growing season.
BIO LANDSLIDES OF THE CINCINNATI, OHIO, AREA
10
t
Ring Widths
Wl &W3
ws ; i
0 60
b=40
W2&W4
1940
Wl +W3
Eccentricity
A A
W4
W2 + W4
1940 1950 1960
YEA R
1970 1980
Figure 7. Thickness of growth rings in the "west tree" of figure 6, plotted as a function of time. Cores are from the uphill (W1), downhill
(W3), and lateral (W2 and W4) sides of the tree. Also plotted is the percent eccentricity of rings on opposite sides of the tree.
We were able to measure a small amount of movement
in an inclinometer casing in boring 8 and with an extensom-
eter (for locations, see pi. 1) during 1979-80. The inclinom-
eter casing was deformed slightly by movement during the
winter of 1979-80. Extensometer 2 extended through a
storm drain that crossed under Delhi Pike about 25 m east of
the pine trees. The upslope end of the extensometer was
attached to a box culvert in nonmoving ground between the
thin landslides uphill and the cracks in Delhi Pike downhill.
Movement was measured from December 1979 to June
1980. The extensometer automatically recorded elongation
in increments of 1.73 mm by punching a paper tape; move-
ment was sampled every 15 min. The first indication of
movement was on January 30,1980, when 5.2 mm was mea-
sured over an 8-hr period. Additional movement of at least 7
mm was distributed over the next 4 months; the recorder was
inoperable for 2!/2 weeks in April. Overall, 12.1 mm was
measured during the 6-month interval.
Although our measurements clearly showed movement
during 1979 and 1980, the tree rings formed in those years
had no irregularities in growth-ring thicknesses (figs. 7 and
8). We suspect that a small amount of movement occurs dur-
ing most years, but large, abrupt movements of the thick
slides at Delhi Pike apparently occurred only during
1958-59 and 1973.
The current owner of the property, Mr. Schmale, can
testify that the perception of small movements is difficult
without a good reference feature for measurement. Even
though the pine trees along his driveway at 5434 Hillside
Avenue had been strongly tilted by 1972, there was no evi-
dence of distress in his house standing less than 10 m down-
slope from the lower toe bulge shown in figure 5 (Mr. Dale
Schmale, oral commun., March 17, 1981). Active landslid-
ing had already damaged property several hundred meters
east of Mr. Schmale's property. However, it was Mr.
Schmale's impression that the ditch and shoulder of Delhi
Pike were intact and free from cracks in 1972. Abrupt move-
ment in the early spring of 1973 revealed several layers of
patching in Delhi Pike, which would have covered preexist-
ing cracks but proves the existence of pre-1972 movement
uphill from Schmale's house. Because the 100-yr-old house
was virtually free of structural distress at the time of abrupt
landsliding in 1973, it is unlikely that displacement had
occurred on throughgoing landslide failure surfaces then.
The displacement occurring in the Delhi Pike apparently
extended to a toe only 10-15 m downhill and did not involve
the house. At the time of our study (1979-81), we were able
to identify two additional landslide toes downslope from the
house, and movement on those failure surfaces responsible
for the landslide toes was measured in inclinometer casings
in borings 9 and 20.
We interpret these observations to mean that the slope
was failing progressively, perhaps beginning with the con-
struction of Delhi Pike in the 1800's. The failure process
consisted of creating multiple failure surfaces, which pro-
duced several toes of movement at increasing distances
downslope, beginning from a single scarp in Delhi Pike. The
100-yr-old farm house was not destroyed until progressive
failure produced a failure surface that emerged farther down
the slope. The single scarp on Delhi Pike is now the upper
part of a landslide that contains at least four toes and, pre-
sumably, four separate failure surfaces that emanate from the
single failure surface near the scarp.
DEFORMATION CAUSED BY THICK LANDSLIDES
One of the more interesting aspects of the deeper seated
landsliding along Hillside Avenue is that so much evidence
LANDSLIDES IN COLLUVIUM fill
10
tn
F .10
10
fc
E1 &E3
Ring Widths
E2&E4
E6Top
Downslope
1959
pith
1940 1950 1960
YEA R
1970 1980
!40
0 60
140
1940
E3
Eccentricity
E1 +E3
E4
E2 + E4
Downslope
U p + Downslope
1 9 5 0 1 9 6 0
Y E A R
1 9 7 0 1 9 8 0
Figure 8. Thickness of growth rings in the "east tree" of figure 6, plotted as a function of time. Cores are from the uphill (El), downhill
(E3), and lateral (E2 and E4) sides of the tree. E6 is the tree core from just above the sharp bend shown in figure 6B. The bend probably
formed in 1958. Also plotted is the percent eccentricity of rings on opposite sides of the tree.
of past development remains in place. We counted nearly 30
foundations along Hillside Avenue between Anderson Ferry
Road and Delhi Pike for houses that apparently have been
destroyed by landsliding. The landslides have been
damaging structures in this area for more than a century.
Homes involved in movement episodes typically are dis-
tressed to the point that they must be razed. Yet, the cumula-
tive movement of all these events has not markedly deflected
the alignment of Hillside Avenue, and foundations of
destroyed structures are recognizable and persist. Evidently,
the deep-seated landslides undergo significant internal
deformation relative to the amount of sliding displacement.
The deformation that is typical in landslide toes is illus-
trated in figure 9. The view is of a landslide toe encroaching
on asphalt pavement along a break in slope at Anderson
Ferry Road and River Road (fig. 1). This photograph, taken
in August 1982, documents deformation that occurred over
a period of about 9 years.
The bulge in the asphalt was about 26 m wide at the
uphill edge and encroached 6.1 m into the blacktopped area.
At the uphill edge, the blacktop was elevated 1.5 m, while it
was displaced horizontally a maximum of 1 m in the middle.
The horizontal displacement diminished to zero at the edges
of the bulge. The distal edge of the bulging blacktop is not
thrust over adjacent, unfailed blacktop but merely elevated
vertically. Thus, the direction of displacement at the leading
and flanking edges of the deformed blacktop in the toe
appears to be nearly vertical upward. Six meters back from
the leading edge, motion is 1.5 vertical to 1.0 horizontal.
Figure 10 shows another example of deformation pro-
duced by movement of a landslide toe. In this case, at a land-
slide in the Mt. Adams area of Cincinnati, the curb of the
street has not been displaced, but two slabs of concrete side-
walk have been lifted and displaced over the curb. For the
larger slab on the right, horizontal displacement is 0.60 m,
and the uphill edge of the slab is elevated 0.82 m.
The early stages of formation of a landslide toe were
expressed along a driveway from Hillside Avenue just west
of Delhi Pike. A slope underlain by colluvium was cut to
provide access for construction of new apartments. Shortly
after the cut was made, cracks formed in the colluvial slope
about 25 m uphill from the edge of the road. A small bulge
formed along the edge of the road, involving as much as 2 m
of the paved surface. A photograph taken in 1982 (fig. 11)
shows a bulged patch on the blacktop surface and a bulge
along the edge of the driveway. In the foreground along this
bulge on the edge of the driveway, several near-vertical
shear surfaces are exposed. Tabular limestone fragments
Figure 9. Bulge produced by landslide toe in asphalt parking area at Anderson Ferry Road and River Road. Vertical displacement in the
toe greatly exceeds horizontal displacement.
Figure 1 0. Landslide toe formed in Mount Adams section of Cincinnati. Ratio of vertical to horizontal displacement is 1.4: 1.
within the colluvium are preferentially oriented with their
minimum dimension normal to the shear surfaces. At other
places, limestone fragments in colluvium do not show a
strong preferential orientation. Here, the fragments have
been reoriented by pervasive shearing such that the cross-
sectional area exposed to shear is a minimum.
SUBSURFACE WATER IN THICK LANDSLIDES
The abrupt movement of thick landslides in April 1973,
which destroyed several homes, occurred during a period of
excessive precipitation. However, over a period of years,
movement of the thick landslides appears to have been
Figure 1 1 . Near-vertical bump formed in landslide toe. Limestone clasts have been preferentially oriented with their minimum dimension
at right angles to the direction of pervasive shearing. The bump has formed without appreciable horizontal offset.
caused as much by construction activity as by excess precip-
itation.
Our drilling and trenching failed to find saturated con-
ditions anywhere within the thick colluvium except in boring
20 at the toe. Saturated conditions must exist at other places
in the colluvium, for we noted a line of springs near boring
12 (pi. 1) and an ephemeral spring near boring 14. Boring 12
was flooded with surface water from the springs and pro-
vided no worthwhile data, but colluvium in borings 9,10,13,
and 14 was unsaturated.
The typical subsurface water condition was illustrated
by measurements in boring 13. There, the colluvium was not
saturated. Free water was first encountered in the first lime-
stone bed under the colluvium that was penetrated by drill-
ing. This water was under artesian pressure and responded to
an annual cycle of maximum pressure in the late spring and
minimum pressure in the early fall (fig. 12). In those parts of
the landslide that extend nearly to the contact between bed-
rock and colluvium (that is, the head region), the artesian
pressures could significantly reduce stability. Most of the
toes of the thick landslides are entirely within the colluvium
and significantly removed from the bedrock containing
artesian water. The effect of the artesian pressure is
dissipated a short distance from the contact between bedrock
and colluvium (R.L. Baum, U.S. Geological Survey, oral
commun., 1985).
Probably during exceptional periods of precipitation,
saturation and elevated pore-water pressures in the thick col-
luvium are more extensive than we found. However, inas-
much as we measured movement of the thick landslide
during a period of below-average precipitation and during a
period when water levels were generally below the various
failure surfaces, saturation of the colluvium is not necessary
for renewed movement on the existing failure surfaces.
SUMMARY OF THICK LANDSLIDES
The thick landslides occur in colluvium that is thicker
than about 2 m. Morphologically, they consist of a simple
head, a complex of toes, and poorly expressed flanks. The
heads occupy a narrow zone containing a few scarps and
open cracks; the horizontal component of displacement in
the heads is smaller than the vertical component. The toes of
movement typically emerge at several places downslope. If
the toe emerges at a sharp break in slope, such as a roadcut,
the vertical component of displacement is much larger than
Boring 13
C olluvium
_Bedrock
(shale)
^Bentonite seal
Open-tube
piezometer
(Limestone)
Zone of
U plift Pressure
9
1979
11 1
TIME (MON THS)
Figure 1 2. Water-level record of one year for an open-tube piezometer (No. 13, pi. 1) sealed in a limestone layer in bedrock. Depth to
layer was 3.4 m, and water level varied from 3.3 m during September 1979 to 1.8 m about May 1, 1980. The line of intersection of the
limestone layer with the base of the colluvium is a zone of artesian pressure that develops a head as great as 1.5 m during the spring. (See
inset figure.) (From Fleming and others, 1981.)
the horizontal component. If the toe emerges at a smooth or
flat part of the slope, displacement is virtually all horizontal,
and the lobe of moving ground may override the preexisting
ground surface. The flanks of the landslide lobes are poorly
expressed and could not be located in the field. Shear dis-
placement on the landslide flanks is evidently distributed
over too broad an area to produce cracking or faulting. How-
ever, the distributed shear may contribute to the extensive
structural damage to homes that seems to take place when
only small amounts of displacement have occurred.
The presence of a single head and multiple toes implies
a single failure surface under the upper part of the landslide
that splays into multiple failure surfaces farther downslope.
The failure surface(s) appear to be contained within the
colluvium rather than at the contact between colluvium and
bedrock.
Eccentric growth rings and reaction wood in tree cores
and multiple generations of asphalt patching on Delhi Pike
are evidence that the thick landslide began to form at the
head and enlarged progressively. The position of the upper-
most cracks in the head roughly coincides with the boundary
between cut and fill in the roadbed. The abrupt movement
that occurred in 1958-59 apparently extended only a few
meters downslope from the cracks in the road to a complex
of toes. The house and other structures on the property were
destroyed by another episode of abrupt movement in 1973.
Presumably, new failure surfaces formed that successively
emerged farther down the slope in 1973 and later. Move-
ment appeared to be continuing in 1990 on at least the deep-
est of these surfaces, which emerged as a bump in new
asphalt on Hillside Avenue.
Abnormally large amounts of annual precipitation coin-
cided with the abrupt movement episodes in 1958 and 1973.
The association between excessive precipitation and lands-
liding has been so thoroughly documented for landslides
around the world that it is unnecessary to do so here. We
measured about 12 mm of movement of the thick Delhi Pike
landslide during our only period of monitoring in 1979-81.
Precipitation during the entire period of our study was aver-
age to below average, and water levels in open piezometers
were generally below the level of the failure surface(s).
Our measurements in open-tube piezometers failed to
document saturated conditions within the colluvial body of
the landslide, although we did find water under artesian pres-
sure in the limestone layers in the underlying bedrock. The
pressures varied on an annual cycle, producing minimum
values during the fall and maximum values in late spring.
Most of the failure surfaces are well above the contact
between bedrock and colluvium and, thus, subsurface water
pressure in the bedrock is not a major control on the reacti-
vated movement of these thick landslides.
THIN LANDSLIDES
The thin landslides occur on the uphill side of Delhi
Pike (fig. 2) where slopes are significantly steeper than in the
area of thick slides. Slope inclinations range from about 18
to 25. An open woods of maple and ash trees, typically
spaced about 3 m apart and lacking any appreciable under-
story, occupies the upper part of the slope. Farther down-
slope, near Delhi Pike, the hillslope supports a dense
understory of various shrubs and vines and a canopy of small
locust and elm trees. The change in vegetation occurs a short
distance from the contact between the Kope and Fairview
Formations.
For the most part, the slope is devoid of obvious scarps,
flanks, or landslide toes. Rather, the hillslope has a very gen-
tle undulatory surface superimposed on a series of irregu-
larly spaced topographic benches. Slopes that have not failed
within historical time may lack the well-defined scarps and
flanks that are commonly observed in landslide deposits.
More recent slope failures show all the classic features of
landslides and, even though they quickly reestablish a vege-
tative cover, they remain recognizable topographic forms.
Our experience with excavations and landslides throughout
the metropolitan area is that virtually all these colluvial
slopes have failed at some time in the past, and a trench any-
where through the colluvium would reveal a continuous fail-
ure surface near the colluvium-bedrock contact.
The typical form of the area underlain by thin landslides
is best expressed toward the western edge of the area shown
in figure 2, uphill from Hillside Avenue. Uppermost scarps
of the slides extend into and beyond areas underlain by Fair-
view Formation and form a crudely scalloped upper bound-
ary within the colluvium. The upslope ends of the landslides
are broad amphitheaters with generally smooth slopes. Over
most of the area, the scarps are expressed as a slight but con-
tinuous break in slope. The portion of the landslide on the
Fairview Formation contains widely scattered benches and
ramps that are subparallel to the topographic contours. These
benches appear to be remnants of slide blocks that have
detached from the scarp and been smoothed over time.
Downslope, on the Kope Formation, the benches and
ramps are more consistently parallel to the topographic con-
tours. The amphitheaters that produce the scalloped upper
limit to sliding exist upslope from the Kope Formation;
downslope, on the Kope Formation, slopes become straight
without significant bulges or swales parallel to the contours.
The benches on these straight slopes are 15-30 m long and
consist of a stretch of slope about 2 m wide that is flatter than
the overall slope and a similar width that is steeper.
Such benches are common on slopes formed on the
upper part of the Kope Formation throughout the Cincinnati
metropolitan area. We initially believed that these features
were toes or lobes of thin landslides (Fleming and others,
1981, fig. 6). Later trenching of several of these benches
revealed that they overlie a limestone layer in the bedrock.
The benches invariably contain a small-displacement listric
fault in the colluvium that connects with the failure surface
at the base of the landslide. Thus, the more or less continuous
benches are not lobes or toes of movement. The listric fault
is evidence of stretching and arching of the colluvium over a
buried limestone layer.
The toe of thin-landslide movement (fig. 2) is at Delhi
Pike where colluvium has spilled onto the uphill side of the
road. Figure 13, a photograph of Delhi Pike from about the
location of trench 2 (for location, see pi. 1), gives the impres-
sion that the road is very crooked. The road, however, con-
tains only a gentle curve toward the left in the middle of the
view and a curve to the right in the distance; otherwise it is
straight. The uphill side of the road is covered in varying
amounts by lobes of thin slide debris that have accumulated
since the road was closed in 1973. The downhill edge of the
road has been displaced by scarps of the thick landslides, but
the uphill side of Delhi Pike has not moved.
The general form of the thin landslides, then, is multiple
scarps and a simple, single toe. This general form is
Figure 1 3. View uphill along Delhi Pike from the location of
trench 2. Landslides affect both sides of road. Despite appearances,
the road is not crooked; thin landslides from upslope have spilled
onto the left side and scarps of thick landslides have broken the
right side of the road.
expressed even more definitively at other thin landslides in
the Cincinnati metropolitan area. Two examples are
described below:
In 1980, a thin landslide on a slope near the Delhi Pike
landslide complex literally pulled itself apart by stretching
movement. The landslide, described by Riestenberg and
Sovonick-Dunford (1983), was broken into numerous sepa-
rate blocks. Some blocks slid completely out of the landslide
scar, and others were connected to each other only by tree
roots. A careful analysis of striations on the exposed failure
surface and a match of trees and roots allowed Riestenberg
and Sovonick-Dunford (1983) to reconstruct the sequence of
failure. The landslide apparently failed from the top down;
two elements at the extreme upper scarp were the first to fail.
Twelve individual blocks could be traced from their initial to
final positions. At the conclusion of movement, several
blocks from upslope overrode blocks that were farther
downslope.
This style of failure (that is, landslide debris being
pulled apart into separate blocks, some of which slide com-
pletely out of the landslide scar) is evidence for a large
reduction in resistance to sliding following the initial move-
ment. Riestenberg and Sovonick-Dunford (1983) attributed
the large initial strength to tree roots. Failure of the roots that
extended through the potential failure surface could have
resulted in a dramatic loss in strength. Riestenberg and
Sovonick-Dunford (1983) calculated that the roots could
have increased the factor of safety against sliding ninefold.
Another example of extreme extension of thin land-
slides is shown in figure 14. This landslide, which occurred
in 1973 a few kilometers east of downtown Cincinnati along
Columbia Parkway, is typical of many thin landslides. The
landslide slid completely out of its site of failure, moved
about 20 m over the ground surface, and spilled over a retain-
ing wall onto Columbia Parkway. In figure 144, part of the
landslide path is indicated by the light-colored material in
the middle of the view. The initial failure occurred at the
uphill end of the light-colored area. Individual landslide
blocks that have separated during movement but remain on
the slope are on the right side of the light-colored area.
Unfailed material is on the left. At the extreme upper end of
the scar, bedrock was exposed (not visible in the photograph)
at the sole of the landslide. The lower part of the landslide
scar contains four discontinuous benches or undulations. At
least two of these benches continue to the left into the
unfailed colluvium and are probably expressions of buried
limestone layers in the bedrock underlying the colluvium.
The benches that are visible in the light-colored area in
figure 14A are formed within the colluvium. Here, the collu-
vium has not failed to the level of the bedrock as part of the
initial landslide but, rather, has been overridden by landslide
debris from the deeper scar upslope. The landslide appar-
ently pushed away a layer of the upper part of the slope
including the vegetation, the pedologic A horizon, and
underlying colluvium to a depth that contained most of the
tree roots. Figure 145 shows a portion of the landslide debris
that has spilled over the retaining wall onto Columbia Park-
way. The retaining wall was not damaged by the shallow
landslide, which, by the time it reached the wall, was moving
over unfailed colluvium. Note the colluvium piled against
the uphill side of the tree on the far left side of figures 14A
and B. This movement history also is suggestive of a large
drop in strength following initial failure of the landslide.
THE THIN LANDSLIDE AT DELHI PIKE
We studied one of the thin landslides uphill from Delhi
Pike that appeared to be actively moving. Borings, trenches,
open-tube piezometers, and movement indicators were
placed as shown in plate 1. Seven shallow, open-tube
piezometers were installed at the bedrock-colluvium inter-
face. One inclinometer was placed in the 6.5-m-deep boring
7. A recording extensometer was installed across the two
scarps near the head of the slides. Trenches 2, 3, and 4 were
excavated through the colluvium into the bedrock to expose
the failure surface. Trench 4 was excavated after a 1-yr
period of monitoring.
This apparently active landslide has well-defined fea-
tures (pi. 1). It is about 65 m long and 30 m wide in its upper
part. The width is indeterminate in the lower part because the
flanks are indistinct. In the upper part of the landslide, the
ground surface is very irregular and bumpy; the bumps there
correspond to individual landslide blocks that have pulled
apart. Depressions separating these blocks are marked by
distinct scarps on their upslope sides. Large flagstones of
limestone (as large as 1.0x0.5x0.15 m) are scattered over the
ground surface, and chips of weathered shale are present in
the depressions between blocks. Topographically, the land-
slide surface is depressed as much as 0.5 m within the flanks
in the upper 40 m. Farther downslope, the flanks are level
with adjacent materials, and the surface contains benches
that are similar to those farther west.
The uppermost scarp, at an approximate elevation of
220 m (720 ft) is about 1 m high and devoid of vegetation.
There are several other scarps in the upper half of the land-
slide. Along some of these scarps, the separation between
landslide blocks is wider at the flanks (2-3 m) than in the
interior of the landslide (<1 m). This pattern of opening
gives the curious impression that displacement was larger at
the flanks than in the interior of the landslide.
In three places uphill and downhill from boring P-5
and uphill from boring P-7 the wavy pattern on the map
Figure 1 4 (on facing page). Thin landslide along Columbia Park-
way, east of downtown Cincinnati. A, Slide path was created when
landslide from upslope slid completely out of its original position
and pushed the colluvium and vegetation, including roots, off the
remainder of the underlying colluvium. Undulations in the light-
colored colluvium in the path of movement are benches reflecting
limestone layers in underlying bedrock. B, Landslide debris has
spilled over retaining wall onto Columbia Parkway.
in plate 1 depicts areas where displacement has separated
the blocks all the way to the failure surface at the collu-
vium-bedrock contact.
An old, abandoned road that crossed the slide a few
meters downhill from extensometer 1 has been offset about
7 m. Farther downhill, an additional 2-3 m of displace-
ment is revealed by the size of scarps. Thus, the maximum
displacement of the landslide was apparently about 10m.
There is no evidence at the toe for 10 m of displace-
ment, and there are no indications of shortening in the land-
slide uphill from the mapped toe at Delhi Pike. Thus, we
suspect that this landslide occurred before the abrupt move-
ment in 1973 that caused closure of Delhi Pike. If so, as
much as 600 m3 of landslide debris may have been hauled
away from the road surface. If not, there is a very large dis-
crepancy between the amount of displacement evident in the
area of stretching and the amount observable at the toe.
No movement was recorded by the extensometer and
the inclinometer during the year of monitoring in 1979-80.
In 1990, the landslide appeared superficially to be
unchanged since the end of our monitoring in 1981; clearly,
it had made no large, abrupt movements in that time. How-
ever, the inclinometer in boring 7 was found to be squeezed
closed at a depth of 2 m. Thus, the thin landslides do displace
colluvium with small creeplike movements as well as the
abrupt stretching failures observed in other places.
The time of activity of thin landslides throughout the
metropolitan area is restricted to the interval between the
spring thaw and the leafing of vegetation. Episodes of move-
ment of thin landslides usually occur during major storms in
the period of mid-March to early May. During the remainder
of the year, either the vegetation is serving as an effective
agent for dewatering the thin colluvium or the ground is fro-
zen. Open-tube piezometers were dry except during the
spring, when water levels responded almost immediately to
rainfall (fig. 15). The behavior of subsurface water in the thin
colluvium has been described by Gokce (1989) and by
Haneberg and Gokce (in press).
C olluvium
Highest
water
Lowest
water level-si;_
'Bentonite
plug
0 .5
Open-tube
piezometer
(in sand)
1.0
Weathered
Bedrock (shale)
12
3/31/80
TIME (ON E-HOU R U N ITS)
Figure 1 5. Water-level and precipitation record for an open-tube piezometer (P-5, pi. 1) placed at contact between colluvium and bedrock.
This piezometer was dry for most of the year but, during the spring, water levels changed dramatically in response to rainfall. This particular
record shows water-level response to a 1.7-cm rainfall on March 30-31,1980. Sketch to the left of the water record shows location of pi-
ezometer relative to colluvium and weathered bedrock. (From Fleming and others, 1981.)
TRENCHES AND THE FAILURE SURFACES
All three trenches into the thin landslides (trenches 2,3,
and 4 on pi. 1) extended through the colluvium and into bed-
rock. There was no evidence that sliding extended into the
bedrock, but virtually all the colluvium overlying the bed-
rock has failed. In all the trenches, one or more failure sur-
faces were present in the colluvium and (or) in the thin
transitional zone to bedrock. Trenches 2 and 3 were short
trenches into the toes of the thin landslides. Trench 4
extended nearly 60 m from a topographic bench near boring
P-5 to Delhi Pike.
The colluvium is separable into two units on the basis
of color. At the surface, the colluvium is a brown to dark-
brown, stony, silty clay. The dark clay overlies a yellow-
brown silty clay. The underlying bedrock is a mottled
olive-gray shale with a few scattered beds of limestone in
the shale. The thin transitional zone between the colluvium
and bedrock contains abundant small chips of shale in a
clayey matrix.
A few open tension cracks were found in the colluvium.
Some of these cracks occurred at the edges of benches,
where arching of the colluvium over the edges would pro-
duce stretching.
The positions and shapes of the failure surfaces are
simple where the landslide geometry is simple and complex
where large stones or bedrock ledges interfere with the slid-
ing. Where the slope is smooth (for example, between 25
and 35 m on the trench 4 profile on pi. 1), the failure sur-
face is a simple planar surface at the contact with weath-
ered bedrock. A bumpy or hummocky ground surface
typically coincides with more complex failure surfaces. A
bench in the bedrock abruptly steepens the slope of the fail-
ure surface at the colluvium-bedrock contact. One to two
meters downslope from the bedrock bench, one or more lis-
tric faults intersect the basal shear surface (as seen, for
example, at 44 and 47 m on the trench 4 profile). The listric
faults, in part, accommodate the change in slope of the
basal failure surface, where it passes over the ledge, and
also allow stretching of the landslide.
These landslides apparently undergo a significant
amount of internal deformation as they move. The flag-
stones of limestone are haphazardly oriented near the
ground surface. With increasing depth, the large dimen-
sions of the stones become parallel to the failure surface.
Voids are under or upslope from some of the stones and, in
a few places, we noted that a layer of soft clay had been
deposited in the void space.
The failure surfaces are typically paper thin, glossy, and
abundantly striated. At all scales of observation, the surfaces
show roughness. Roughness was produced by small stones
or fossil fragments projecting into the surface. Small steps,
2-3 mm high, are common where the colluvium breaks from
one shiny surface to another. These steps have the same
appearance as chatter marks on fault surfaces in rock.
At a slightly larger scale, the failure surfaces commonly
bifurcate around rock fragments. Also, multiple surfaces
occur where the slope of the failure surface is changing over
a short distance. Still larger undulations in the failure sur-
face occur where a limestone layer produces a bench on the
top of the bedrock. When multiple failure surfaces are
present, as in trench 3 (pi. 1), the material between the sur-
faces is softer and more moist than that outside the failed
zone.
Figure 16 shows two scanning electron micrographs of
fragments of the failure surface. Figure 16A, magnified at
16x, is a view normal to a major failure surface. Direction
of sliding was from right to left. For the most part, the sur-
face is streamlined in the direction of sliding and contains
discontinuous ridges and furrows parallel to the direction of
sliding.
The light-gray circular spots on the failure surface that
are about 0.2 mm in diameter are aggregates of calcium sul-
fate crystals. We observed clear water on the surfaces of soil
peds from the pedologic B horizon but did not directly
observe water along the failure surface. The crystal aggre-
gates are indirect evidence that free water exists seasonally
along the surface of sliding. The crystal aggregates were not
found in the interiors of the blocks or on smaller, subordinate
shiny surfaces such as the one shown in figure 16B.
Figure 16B is an oblique view, at about 40x magnifica-
tion, of a short slickensided surface in a sample containing
several slickensided surfaces. The left and right sides of the
sample (out of view in this image) are both slickensided sur-
faces. The flat, slickensided surface in the middle of the view
does not visibly extend through the sample, which is only
about 1 cm thick. The edge of another smooth surface, sub-
parallel to the first, is exposed about 150 |im to the left of it.
None of these surfaces appear to be parallel to the others, but
neither are they grossly out of alignment with each other, as
might be expected if they were conjugate shear fractures.
Our supposition is that the complex fracture zones occur
where the failure surface is changing shape either by passing
over an asperity or emerging to the ground surface. These
particular samples are from near the downslope end of trench
4, where shortening deformation was occurring and material
was pushed onto Delhi Pike.
The continuous, well-developed failure surface trace-
able in trench 4 (pi. 1, profile) indicates that the strength is at
a residual value (Skempton, 1964, 1985). However, three
other observations suggest that the strength mobilized in the
field might be larger than the laboratory-measured residual
strength. One observation, already mentioned, is roughness
on the failure surfaces that could contribute to increased
strength. The roughness was observed over a scale of about
six orders of magnitude, from micrometers to meters. We
also noted that sliding is accompanied by deformation of the
colluvium over a zone that extends to about a half meter
above the failure surface. Large limestone clasts within this
zone are preferentially oriented parallel to the failure
Figure 1 6. Scanning-electron micrographs of portions of the failure surface. A, View normal to a slickensided surface. B, View oblique
to several subparallel slickensided surfaces.
surface, whereas clasts outside this zone appear to be hap-
hazardly oriented. The deformation that creates parallel
clasts is work done by the landslide in the process of sliding
and contributes to resistance to sliding. And, finally, the
presence of multiple failure surfaces is evidence that the fail-
ure geometry is less than optimum for sliding. Presumably,
multiple slip surfaces are also an indication of internal defor-
mation of landslide material.
We do not know if the internal remolding of material
and the creation of multiple slickensided surfaces occur
simultaneously during landslide movement. If they do, the
residual strength of the soil, as back-calculated from field
conditions, might be significantly larger than the strength
measured on a single failure surface in the laboratory.
There is evidence that landslide movement may be lim-
ited to only one of the slickensided surfaces during a move-
ment episode. The roots that extended into bedrock near the
downslope end of trench 4 were kinked at only one level
even though they penetrated at least two slickensided sur-
faces (fig. 17). Perhaps multiple failure surfaces are created
during the first-time failure of the slope and subsequent
movement involves a more simple geometry.
PROPERTIES OF THE COLLUVIUM
The colluvium is a stony, silty clay containing variable
amounts of sand to boulder-size fragments of limestone.
Within the colluvium, soil scientists recognize two soil
series whose distributions roughly correspond to areas
Figure 1 7. Kinked roots on an exposed failure surface. The
abrupt kinks in the roots formed at the position of the failure sur-
face. There were at least two failure surfaces at this position near
the toe of the landslide, but the roots were deformed only along the
surface exposed in the photographs. (Photograph by Mary Riesten-
berg, College of Mount St. Joseph.)
containing thin and thick landslides (Lerch and others,
1982). These soils are named the Pate Series, representative
of thick colluvium, and the Eden Series, representative of
thin colluvium. In general, the colluvium consists of dark-
brown silty clay (10YR4/3) overlying yellow-brown silty
Table 1 . Summary of selected physical properties of Eden and Pate Soil Series
B21
Data from Soil Conservation Service (1978)
Soil
series
Eden
Pate
Soil
series
Pate
Depth
(cm)
0-15
15-61
0-15
15-91
91-183
Depth
(cm)
0-30
30-95
95-144
Soil T . . .
, . Liquid
hon- ,. ..
limit
zon
A 35-65
B 45-75
A 25-35
Bt 40-60
B3 40-60
Soil
horizon
A
Bt 1
B3
Plas-
ticity
index
12-25
20-45
8-20
15-30
20-35
Data
Permea-
bility
(um/s) j
0.4-4
0.4-1
1 -4
0.4
0.4
1
1
1
1
1
from Norton
Moist
bulk (
density
(g/cm3 )
.35-1.65
.45-1.65
.35-1.60
.50-1.70
.60-1.80
(1979)
Gradation (percent)
Coarse (>76 um)
Gravel
0-40
10-45
0-10
0-10
15-50
Sand
0-25
0-35
0-20
0-20
0-30
Composition of clay minerals in clay-size fraction (percent)
Vermiculite
9
30
34
Micas and illite
77
47
48
Kaolinite
15
23
17
Fine (< 76 um)
Silt
0-30
5-35
5-35
5-30
10-35
Clay
27-60
40-60
20-40
35-55
35-55
Clay fraction
(percent)
23
31
20
1 The Bt horizons contain 2-5 percent montmorillonite. Other horizons contain only trace amounts of
montmorillonite.
clay (10YR4/4). Within a given soil profile, the thicknesses
of both the dark and yellow-brown coliuvium are highly
variable. And, although the dark coliuvium in most places
occurs near the ground surface, it occurs locally as wedges
or layers within the yellow-brown coliuvium. The burial of
surficial layers may be due in some cases to root throw from
overturning of trees (Hack and Goodlett, 1960) and in other
cases to the overriding of surficial materials by landslides.
We noted overridden soil in trench 1 (pi. 1). The pedologic
B horizon contains translocated clay and has a prismatic
blocky structure. Free water was observed on the surfaces of
soil peds in the B horizon shortly after rain storms.
Other than small variations in color and texture, the
physical properties are the same for both soil series. The
routine engineering properties of the coliuvium are reported
in tables 1 and 2. Data in table 1 are from the Soil Conser-
vation Service (SCS) (1977, 1978) and Norton (1979) for
the entire range of the Pate and Eden Soil Series and thus
are representative of the coliuvium over a wide geographi-
cal area.
The principal difference in properties between the Pate
and Eden Series is in the amount of coarse fraction. Eden
soils, which occur on the higher, steeper reaches of the
slopes, contain more rock fragments than the Pate soils. The
amount of increase is illustrated by the increase in rock frag-
ments larger than about 10 cm exposed in the walls of the
Table 2. Summary of physical properties of coliuvium and shale
from the Delhi Pike study area
[All tests performed according to American Society for Testing and Materials stan-
dards]
Physical property Average Range
Atterburg limits (20 tests):
Liquid limit ....................... 45 36-52
Plastic limit ....................... 23 19-26
Plasticity index .................. 22 15-27
Gradation (20 tests; percent):
Sand .................................. 10 4-18
Silt .................................... 32 22-44
Clay................................... 59 47-68
Permeability (8 tests; cm/s)* ..... 1.4xlO~3 4xlO~3-
Field density (9 tests; g/cm3 ) .... 2.01 1.74-2.13
Specific gravity of solids (1 test) 2.79
*Nine samples tested, undisturbed in the trench; data from one
sample (SxlO"6) not included in range or average.
trenches. In trench 1, the most downhill trench, rock frag-
ments occupied just 1.4 percent of the exposed trench wall.
In trenches 2 and 3, just uphill from Delhi Pike, rock frag-
ments compose 3.5 and 2.6 percent, respectively, of the
trench walls. In trench 4,6-12 m farther uphill than trenches
B22
2 and 3, rock fragments compose 4.6 percent of the trench
wall. And, at the upper end of trench 4, rock fragments
compose 9 percent of the trench wall. While these amounts
of coarse fraction might influence the physical properties of
the colluvium somewhat, nowhere is the coarse fraction
large enough to dominate the behavior of the soil. The
mechanical behavior of the soil is dominated by the proper-
ties of the fine fraction.
We tested 20 samples from the borings and trenches
including dark-brown colluvium, yellow-brown colluvium,
and weathered shale for Atterberg limits, gradation, per-
meability, field density, and specific gravity of the solids.
Results are summarized in table 2. We did not quantify the
composition of clay minerals in the clay-size fraction, but
they are predominantly illite and mixed-layer clays and have
subordinate amounts of kaolinite.
The only significant difference between our data in
table 2 (which are specifically for the Delhi Pike site) and the
SCS data in table 1 (which represent a large geographic area)
is in the values of permeability. We consistently obtained
values more than 10 times larger than those in the SCS data.
These discrepancies are probably due more to differences in
the test methods than to differences in the materials. Our
tested samples were trimmed from block samples obtained
from the trench walls using the falling-head method (Lambe
and Whitman, 1969), whereas the SCS measurements were
based on percolation tests in the field.
The most important physical property of the soil, with
respect to slope stability, is its strength. Among the several
methods used to measure the strength of a soil, engineering
practice has demonstrated that the relevant measure of resis-
tance to sliding is the residual strength (Skempton, 1964).
Residual strength is commonly measured in a direct- or ring-
shear device that accumulates large displacements on a well-
developed failure surface. The residual strength is the mini-
mum strength measured under fully drained conditions for at
least three different normal loads.
We conducted five separate tests of residual strength on
samples collected from the trench walls. Three of the tests
were by repeated direct shear and two by ring shear. One of
the direct-shear tests was on a remolded total sample that
contained coarse sand and fossil fragments. The other two
direct-shear tests and both of the ring-shear tests had various
coarse fractions removed: the +230 mesh (>63 |im) fraction
was removed for one test of each type, and the +30 mesh
(>600 urn) fraction was removed for the other test of each
type. All the tests were conducted under three different nor-
mal loads equivalent to burial at depths of 1,2, and 4 m. The
results of the tests are in table 3.
Residual strength parameters, expressed as the slope of
the best-fit line through the three data points, are angles of
internal friction ranging between 12 and 24. The largest
value was obtained for the sample that did not have the
coarse fraction removed. Smallest values, 12-17, were
consistently obtained with the ring-shear device on samples
Table 3. Residual strength of colluvium from the Delhi Pike
study area
[All tests performed according to American Society for Testing and Materials stan-
dards]
Strength parameters
Test conditions Residual
friction
(degrees)
Residual
cohesion
(kPa)
Direct-shear test
Total sample:
Cohesion assumed zero .
"Best fit" ......................
+30-mesh fraction removed:
"Best fit" ......................
Minimal friction ...........
"Best fit" ......................
Minimal friction ...........
24
21
21
18.5
12
21
20
11
0
3
0
8
16
0
5
16
Ring-shear test
"Best fit" ......................
"Best fit" ......................
17.5
12
16
12.5
0
5.5
0
3
that had the coarse fraction removed. Whether the fraction
removed was +30 mesh or +230 mesh made only a slight dif-
ference in the strength parameters.
The best-fit straight line through the data points for
each test had a small cohesion intercept that ranged from 3
to 16 kPa. There has been much discussion in the literature
about whether a sample at residual strength can have a small
cohesion intercept as part of its total strength parameters.
The data obtained here are not relevant to that issue, and
interested readers are referred to Skempton (1985) and to the
references that he cites relative to residual strength.
Shannon and Wilson, Inc. (oral commun., 1980) stud-
ied a large landslide in colluvium in the Mt. Adams section
of Cincinnati (fig. 1). Their laboratory measurements
showed that the colluvium had residual friction angles of 15
and zero cohesion. Back-calculation of strength parameters
for the landslide produced a residual friction angle of 13 and
a cohesion of 96 kPa.
STABILITY ANALYSES
We performed stability analyses for both thin and thick
landslides at the Delhi Pike site. For the thin landslides, we
used the geometry revealed by trench 4 (pi. 1) and averages
B23
of measured properties. Figure 15 shows the response for
water levels in thin slides to even a moderate-size spring
rainstorm. Storms large enough to raise the unconfmed water
surface as high as the ground surface are likely even in years
of average precipitation. We back-calculated strength
parameters for the above conditions using the method of
Janbu (1973). At equilibrium, a frictional resistance of 39 is
necessary to support the slope. This value is much larger
than the strength we would expect to find along the well-
developed failure surfaces exposed in trench 4.
Skempton (1964,1985) reports measurements of resid-
ual strength on clays of similar physical properties in the
range of 13 to 15. Results of our measurements of residual
friction angles of the colluvium varied from about 24 to 12
for different types of tests and methods of sample prepara-
tion (table 3). The largest values were obtained from
untreated but remolded samples tested in a direct-shear
machine. The smallest values were obtained on samples that
had the coarse fraction (>30 and >230 mesh) removed and
were tested in a ring-shear device. After the tests, we
inspected the failure surfaces developed during shearing.
Samples having larger residual strength invariably contained
sand-size particles along the planes of failure.
Even assuming the maximum value of residual strength
that we measured, the colluvium is only marginally strong
enough to resist failure by thin iandsliding. Therefore, some
additional factor(s) must have enhanced the resistance to
sliding. Possible factors mentioned earlier include tree roots,
roughness of the failure surface, and (or) small residual
cohesion.
The factor of safety of slopes underlain by thick land-
slides is apparently consistent with laboratory measurements
of strength. We back-calculated the stability of the slope that
we studied on the downhill side of Delhi Pike, using the
same properties as for the thin slides. The computed effec-
tive residual friction angles necessary for equilibrium range
from 23 for a slope saturated to the ground surface to 12
for a slope with the water level below the failure surface.
Both extreme values appear to be unrealistic. We measured
saturated conditions to a level of 1 m above the failure sur-
face at the toe and noted the presence of springs locally
within the failed mass. On the other hand, near the middle
of the landslide, we did not find saturated conditions in the
colluvium anywhere above the bedrock; the failure surface
in the colluvium was at least 2 m above the bedrock-collu-
vium contact. Thus, an estimated average residual friction in
the range of 16 to 20 would be consistent with field and
laboratory data.
CONCLUDING REMARKS
In summary, slopes in thin colluvium appear to be more
stable than we would predict from laboratory measurements
of residual strength of the colluvium. The puzzling question
0 6
0 .5
0 .2
x-~Thin landslides
Residual strength
Thick landslides
IN C REA SIN G DISPLA C EMEN T -
Figure 1 8. Hypothetical force-ratio-displacement models for thin
and thick landslides. Back-calculation of stability of thin colluvium
reveals that the residual strength is insufficient to support the slope.
Field behavior indicates that some form of stick-slip behavior pro-
duces large initial strength that is lost after a small amount of dis-
placement The back-calculation of stability of thick colluvium
reveals that failure is consistent with the measured value of residual
strength.
about thin landslides, then, is not why some slopes have
failed, but rather why failure is not more extensive. Observa-
tions of episodes of movement of typical thin landslides (fig.
18) suggest that a type of stick-slip phenomenon may con-
tribute to iarger-than-measured initial strength. We
observed that a small displacement causes a large loss of
resistance, and a thin landslide commonly accelerates out of
the slide scar. The situation, shown schematically in figure
18, may be equivalent to the abrupt loss of strength in first-
time failures of slopes compared to reactivation of previ-
ously failed slopes. For thin colluvium, however, the slope
has previously failed and yet retains greater resistance to
sliding than predicted by laboratory measurements.
The landslides in the thick colluvium behave more or
less as we would predict from laboratory measurements of
residual strength of the colluvium. We know of no examples
in which thick landslides have moved more than a few
meters during any given episode of movement. More com-
monly, the total displacement during a movement episode is
only a few centimeters. A force-displacement diagram for
the thick landslides (fig. 18) illustrates the concept that the
maximum mobilized resistance to sliding is residual friction.
The cause of the difference in behavior between the two
landslide types could be due to a small amount of residual
cohesion and/or roughness of the failure surface and perhaps
tree roots.
The presence of multiple failure surfaces is an interest-
ing aspect of the thick landslides. The history of failure at the
Schmale residence indicated that the failure surfaces formed
progressively from a common headscarp along Delhi Pike.
Our stability analysis is relevant only to the deepest of the
failure surfaces. The tree rings and constricted inclinometer
casings are evidence that movement continues to occur on
the more shallow failure surfaces.
REFERENCES CITED
Agard, S.S., 1979, Investigation of recent mass movements near
Telluride, Colorado, using the growth and form of trees: Boul-
der, Colo., University of Colorado, M.S. thesis, 157 p.
Durrell, R.H., 1961a, The Pleistocene geology of the Cincinnati
area, Field Trip 3: Geological Society of America Guidebook
for Field Trips, Cincinnati, 1961, p. 47-57.
1961b, Geomorphology of the Cincinnati area: Geological
Society of America Guidebook for Field Trips, Cincinnati,
1961, p. 219-223.
Earth Surface Process Group, 1987, Report and recommendations
on maintenance and repair of deteriorating retaining walls and
streets damaged by landslides, City of Cincinnati: Unpublished
report of University of Cincinnati/City of Cincinnati Commit-
tee submitted to the Schmale Commission, City of Cincinnati,
50 p. Available from Cincinnati Department of Public Works.
Fleming, R.W., Johnson, A.M., and Hough, I.E., 1981, Engineering
geology of the Cincinnati area [field trip 18], in Roberts, T.G.,
ed., Geomorphology, hydrogeology, geoarcheology, engineer-
ing geology, v. 3 o/GSA Cincinnati '81 Field trip guide-
books: Falls Church, Va., American Geological Institute,
p. 543-570.
Fleming, R.W., and Taylor, F.A., 1980, Estimating the costs of
landslide damage in the United States: U.S. Geological Survey
Circular 832, 21 p.
Ford, J.P., 1974, Bedrock geology of the Cincinnati West quadran-
gle and part of the Covington quadrangle, Hamilton County,
Ohio: Ohio Division of Geological Survey Report of Investi-
gations 93, map with text, scale 1: 24,000.
Gibbons, A.B., 1972, Geologic map of parts of the Burlington and
Addyston quadrangles, Boone County, Kentucky: U.S.
Geological Survey Geologic Quadrangle Map GQ-1025, scale
1: 24,000.
Gokce, A.O., 1989, Mechanisms of landsliding in the Cincinnati
area: University of Cincinnati, Department of Geology, Ph.D.
thesis, 507 p.
Hack, J.T., and Goodlett, J.C., 1960, Geomorphology and forest
ecology of a mountain region in the central Appalachians: U.S.
Geological Survey Professional Paper 347,66 p.
Haneberg, W.C., and Gokce, A.O., in press, Rapid water-level fluc-
tuations in a thin colluvium landslide west of Cincinnati, Ohio:
U.S. Geological Survey Bulletin 2059-C.
Janbu, Nils, 1973, Slope-stability computations, in Hirschfield,
R.C., and Poulos, S.J., eds., Embankment-dam engineering,
Casagrande volume: New York, John Wiley, p. 47-86.
Kozlowski, T.T., 1971, Growth and development of trees, volume
II: New York, Academic Press, 514 p.
Lambe, T.W., and Whitman, R.V., 1969, Soil mechanics: New
York, John Wiley and Sons, Inc., 553 p.
Lerch, N.K., Hale, W.F., and Lemaster, D.D., 1982, Soil survey of
Hamilton County [Ohio]: U.S. Department of Agriculture, Soil
Conservation Service County Soil Survey, 219 p.
Luft, S.J., 1971, Geologic map of part of the Covington quadrangle,
northern Kentucky: U.S. Geological Survey Quadrangle Map
GQ-955, scale 1: 24,000.
1972, Geologic map of the Butler quadrangle, Pendleton
and Campbell Counties, Kentucky: U.S. Geological Survey
Geologic Quadrangle Map GQ-982, scale 1: 24,000.
Norton, L.D., 1979, Soils developed from the Kope geologic for-
mation: Ohio Journal of Science, v. 79, no. 4, p. 160-163.
Osborne, R.H., 1974, Bedrock geology of the Cincinnati East quad-
rangle, Hamilton County, Ohio: Ohio Division of Geological
Survey, Report of Investigations 94, map with text, scale
1: 24,000.
Riestenberg, M.M., and Sovonick-Dunford, Susan, 1983, The role
of woody vegetation in stabilizing slopes in the Cincinnati
area, Ohio: Geological Society of America Bulletin, v. 94,
p. 506-518.
Skempton, A.W., 1964, Long-term stability of clay slopes: Ge*o-
technique, v. 14, p. 77-102.
1985, Residual strength of clays in landslides, folded strata
and the laboratory: Ge*otechnique, v. 35, p. 3-18.
Soil Conservation Service, 1977, Pate Series: U.S. Department of
Agriculture, National Cooperative Soil Survey, Tentative Soil
Series Description AKN-FWS-DWW, 7 p.
1978, Eden Series: U.S. Department of Agriculture,
National Cooperative Soil Survey, Soil Series Description
BCW: JHN: RGH, 4 p.
Von Schlichten, O.C., 1935, Landslides in the vicinity of Cincin-
nati: The Compass, v. 15, no. 3, p. 151-154.
Manuscript approved for publication March 30,1993.
Published in the Central Region, Denver, Colorado.
Photocomposition by Shelly A. Fields.
Graphics by Gayle M. Dumonceaux.
Edited by Peter L. Martin.
Cover design by Arthur L. Isom, based on an original drawing by
K.A. Richards, University of Cinicinnati.
. S . GOVE R NME NT PR INTING OFFICE : 1 9 9 3-5 7 3-0 38 /8 6 0 1 0

Report Colluvium

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Report Colluvium

Uploaded by

Copyright:

Available Formats

LANDSLIDES OF THE CINCINNATI, OHIO, AREA

You might also like